DAL Toolbox Deep Python

daltoolboxdp extends daltoolbox with Python-backed components, with emphasis on deep learning and Python-native modeling. In the package name, dp stands for Deep Python.

It currently focuses on:

• Deep learning models backed by torch
• Scikit-learn classifiers exposed through the daltoolbox API
• Time-series forecasting models backed by Python
• Integration of Python model objects into the daltoolbox architecture

These capabilities rely on the reticulate bridge, so the package can keep the object and workflow conventions of daltoolbox while delegating training, encoding, and prediction to Python libraries such as torch and scikit-learn.

The architecture is inspired by the Experiment Lines approach, which promotes modularity, extensibility, and interoperability across tools.
More information on Experiment Lines is available in Ogasawara et al. (2009).

Examples

Example scripts are available at:

• Classification
• Autoencoders
• Time Series

Installation

You can install the latest stable version from CRAN:

install.packages("daltoolboxdp")

To install the development version from GitHub:

library(devtools)
devtools::install_github("cefet-rj-dal/daltoolboxdp", force = TRUE, dependencies = FALSE, upgrade = "never")

Bug reports and feature requests

Please report issues or suggest new features via:

• GitHub Issues